-
A general joint latent class model of longitudinal and survival data with the covariance modelling
Authors:
Ruoyu Miao,
Christiana Charalambous
Abstract:
Based on the proposed time-varying JLCM (Miao and Charalambous, 2022), the heterogeneous random covariance matrix can also be considered, and a regression submodel for the variance-covariance matrix of the multivariate latent random effects can be added to the joint latent class model. A general joint latent class model with heterogeneous random-effects modelling is a natural extension of the time…
▽ More
Based on the proposed time-varying JLCM (Miao and Charalambous, 2022), the heterogeneous random covariance matrix can also be considered, and a regression submodel for the variance-covariance matrix of the multivariate latent random effects can be added to the joint latent class model. A general joint latent class model with heterogeneous random-effects modelling is a natural extension of the time-varying JLCM, which consists of the linear and the log link functions to model the covariance matrices as the variance-covariance regression submodel based on the modified Cholesky decomposition, longitudinal submodel, survival submodel as well as the membership probability. It can help to get more information from the random covariance matrix through the regression submodel and get unbiased estimates for all parameters by modelling the variance-covariance matrix. By adding the regression model, the homogeneous random effects assumption can be tested and the issue of high-dimensional heterogeneous random effects can be easily solved. The Bayesian approach will be used to estimate the data. DIC value is the criterion for deciding the optimal k value. We illustrate our general JLCM on a real data set of AIDS study and we are interested in the prospective accuracy of our proposed JLCM as well as doing the dynamic predictions for time-to-death in the joint model using the longitudinal CD4 cell count measurements.
△ Less
Submitted 24 March, 2025;
originally announced March 2025.
-
A time-varying bivariate copula joint model for longitudinal and time-to-event data
Authors:
Zili Zhang,
Christiana Charalambous,
Peter Foster
Abstract:
A time-varying bivariate copula joint model, which models the repeatedly measured longitudinal outcome at each time point and the survival data jointly by both the random effects and time-varying bivariate copulas, is proposed in this paper. A regular joint model normally supposes there exist subject-specific latent random effects or classes shared by the longitudinal and time-to-event processes a…
▽ More
A time-varying bivariate copula joint model, which models the repeatedly measured longitudinal outcome at each time point and the survival data jointly by both the random effects and time-varying bivariate copulas, is proposed in this paper. A regular joint model normally supposes there exist subject-specific latent random effects or classes shared by the longitudinal and time-to-event processes and the two processes are conditionally independent given these latent variables. Under this assumption, the joint likelihood of the two processes is straightforward to derive and their association, as well as heterogeneity among the population, are naturally introduced by the unobservable latent variables. However, because of the unobservable nature of these latent variables, the conditional independence assumption is difficult to verify. Therefore, besides the random effects, a time-varying bivariate copula is introduced to account for the extra time-dependent association between the two processes. The proposed model includes a regular joint model as a special case under some copulas. Simulation studies indicates the parameter estimators in the proposed model are robust against copula misspecification and it has superior performance in predicting survival probabilities compared to the regular joint model. A real data application on the Primary biliary cirrhosis (PBC) data is performed.
△ Less
Submitted 30 November, 2024; v1 submitted 11 September, 2022;
originally announced September 2022.
-
A joint latent class model of longitudinal and survival data with a time-varying membership probability
Authors:
Ruoyu Miao,
Christiana Charalambous
Abstract:
Joint latent class modelling has been developed considerably in the past two decades. In some instances, the models are linked by the latent class k (i.e. the number of subgroups), in others they are joined by shared random effects or a heterogeneous random covariance matrix. We propose an extension to the joint latent class model (JLCM) in which probabilities of subjects being in latent class k c…
▽ More
Joint latent class modelling has been developed considerably in the past two decades. In some instances, the models are linked by the latent class k (i.e. the number of subgroups), in others they are joined by shared random effects or a heterogeneous random covariance matrix. We propose an extension to the joint latent class model (JLCM) in which probabilities of subjects being in latent class k can be set to vary with time. This can be a more flexible way to analyse the effect of treatments to patients. For example, a patient may be in period I at the first visit time and may move to period II at the second visit time, implying the treatment the patient had before might be noneffective at the following visit time. For a dataset with these particular features, the joint latent class model which allows jumps among different subgroups can potentially provide more information as well as more accurate estimation and prediction results compared to the basic JLCM. A Bayesian approach is used to do the estimation and a DIC criterion is used to decide the optimal number of classes. Simulation results indicate that the proposed model produces accurate results and the time-varying JLCM outperforms the basic JLCM. We also illustrate the performance of our proposed JLCM on the aids data (Goldman et al., 1996).
△ Less
Submitted 1 March, 2023; v1 submitted 22 June, 2022;
originally announced June 2022.
-
A Gaussian copula joint model for longitudinal and time-to-event data with random effects
Authors:
Zili Zhang,
Christiana Charalambous,
Peter Foster
Abstract:
Longitudinal and survival sub-models are two building blocks for joint modelling of longitudinal and time to event data. Extensive research indicates separate analysis of these two processes could result in biased outputs due to their associations. Conditional independence between measurements of biomarkers and event time process given latent classes or random effects is a common approach for char…
▽ More
Longitudinal and survival sub-models are two building blocks for joint modelling of longitudinal and time to event data. Extensive research indicates separate analysis of these two processes could result in biased outputs due to their associations. Conditional independence between measurements of biomarkers and event time process given latent classes or random effects is a common approach for characterising the association between the two sub-models while taking the heterogeneity among the population into account. However, this assumption is tricky to validate because of the unobservable latent variables. Thus a Gaussian copula joint model with random effects is proposed to accommodate the scenarios where the conditional independence assumption is questionable. In our proposed model, the conventional joint model assuming conditional independence is a special case when the association parameter in the Gaussian copula shrinks to zero. Simulation studies and real data application are carried out to evaluate the performance of our proposed model. In addition, personalised dynamic predictions of survival probabilities are obtained based on the proposed model and comparisons are made to the predictions obtained under the conventional joint model.
△ Less
Submitted 21 September, 2022; v1 submitted 3 December, 2021;
originally announced December 2021.
-
Joint modelling of longitudinal measurements and survival times via a multivariate copula approach
Authors:
Zili Zhang,
Christiana Charalambous,
Peter Foster
Abstract:
Joint modelling of longitudinal and time-to-event data is usually described by a joint model which uses shared or correlated latent effects to capture associations between the two processes. Under this framework, the joint distribution of the two processes can be derived straightforwardly by assuming conditional independence given the random effects. Alternative approaches to induce interdependenc…
▽ More
Joint modelling of longitudinal and time-to-event data is usually described by a joint model which uses shared or correlated latent effects to capture associations between the two processes. Under this framework, the joint distribution of the two processes can be derived straightforwardly by assuming conditional independence given the random effects. Alternative approaches to induce interdependency into sub-models have also been considered in the literature and one such approach is using copulas to introduce non-linear correlation between the marginal distributions of the longitudinal and time-to-event processes. The multivariate Gaussian copula joint model has been proposed in the literature to fit joint data by applying a Monte Carlo expectation-maximisation algorithm. In this paper, we propose an exact likelihood estimation approach to replace the more computationally expensive Monte Carlo expectation-maximisation algorithm and we consider results based on using both the multivariate Gaussian and $t$ copula functions. We also provide a straightforward way to compute dynamic predictions of survival probabilities, showing that our proposed model is comparable in prediction performance to the shared random effects joint model.
△ Less
Submitted 3 March, 2022; v1 submitted 27 August, 2021;
originally announced August 2021.
